Chromatography Forum

Here's a question: In Chiral separations, what is the basic mechanism of separation. I know that the stationary phase is chiral, so you get a diastereomeric interaction, therefore you get slightly different behavior for 2 isomers, bla, bla, etc.

But, aside from this, what is the basic mechanism of the separation. Is it reversed phase...or....?

Thanks
Adam

Here's an answer " all of the above".

Commonly, chiral stationary phases can utilise:- normal phase, polar ionic, polar organic, and reverse phase modes. Some macrocyclic glycopeptides columns ( eg Vancomycin ), can be used in any of the four modes.

The interactions involved can be:- hydrogen bonding, pi-pi, steric, ionic dipole, and inclusion, or any compbination of those interactions.

The most difficult step is to select the column and system that will work for novel samples.

There are several good books written, and there are some good www sites ( eg Daicel, Regis, Astec ), that provide fairly detailed discussions on what their columns can separate.

There are also several threads on this site, if you search on Chiral or CSP.

Bruce Hamilton

Bruce, I like your answer!

To elaborate on it a bit, to a first approximation, chiral separarations are just like any other separation; retention is driven by primarily by hydrophobicity, or adsorption, or ion-exchange, or partition, or . . .

The chiral part of it is a secondary effect that lets one enantiomer be held (or not held) a bit more tightly than the other.

I don't know guys. Is it really that complicated?

Your answers seem to indicate that it is always a mixed mode situation (i.e. a combination of mechanisms). But that makes for a very challenging problem for the analyst.

Can't we have - for example - a reversed phase chiral column. It seems like that would be what most folks would want.

Thanks for your input.

I don't know guys. Is it really that complicated?.

Many of those mechanisms are used by the Vancomycin column that I'm current trying to achieve a chiral separation on. The screening mobile phases are:-
Polar Ionic Mode : 100/0.1/0.1 (v/v/v): MeOH/HOAc/TEA
Reverse Phase : 30/70: MeOH/20mM NH4OAc, pH 5
Polar Organic Mode : 100% EtOH
Normal Phase : 30/70: EtOH/heptane

If any of them work, then I can refine the mobile phase, if not I'm onto another chiral column. There are four-five main categories of chiral stationary phases,

1. Brush type ( Pirkle ) donor acceptors - eg Whelk, (S)-tert-Leucine + (R)--(alpha-naphthyl)ethylamine.
2. Cellulose triesters or carbamates on silica - eg Chiralcel OB
3. Inclusion - eg cyclodextrins, crown ethers, polyacrylates etc.
4. Ligand exchangers - eg proline, where the mobile phase also contains a metal ion.
5. Protein - albumin, glycoprotein.

The diversity exists for a reason. I should note these columns are usually very fragile, and the wrong solvent can kill them. The best solvents for one can be fatal to another.

Derek Lowe put the following on his blog at the beginning of this month:-
" Back some years ago, the guys down the hall from me had bought one of the largest Chiracel columns that were then sold. ....This was one of the ones where the chiral packing wasn't really bonded on to anything, but just sort of layered on another powdered solid support. And as the literature included with the column made clear, this meant that you could wash the stuff right off if you weren't careful with your solvent selection.

Well, it made it clear if you, like, read the sheet and everything. Which didn't stop someone from taking up their compound in methylene chloride and pumping it right onto the barely-used $15,000 (late 1980s money) column. And in the fullness of time (say, ten or fifteen minutes), out came the solvent front from the other end: cloudy, milky, swirling with opalescent shimmers like shampoo. Which shimmery stuff was, of course, the fifteen long ones of chiral resolving agent, scoured off the packing material by the cleansing wave of chlorinated solvent.

There: clean, simple, direct, and easily avoidable by spending two minutes reading a sheet of paper. "

Your answers seem to indicate that it is always a mixed mode situation (i.e. a combination of mechanisms). But that makes for a very challenging problem for the analyst.

Can't we have - for example - a reversed phase chiral column. It seems like that would be what most folks would want.

Some of the interactions are simple, but I still can't estimate what will work with a particular enantiomer. Others might be able to, but I can't.

The problem is that most chiral columns depend on selectivity that is unique to the enantiomer structure. You can estimate whether a column will work by looking at the structure of the enantiomers in the mixture you want to separate, and often the distance of functional groups from the chiral centre can have a significant effect.

However, you have to do the chromatography to confirm. Vancomycin has 18 chiral centres surrounding three inclusion cavities - my meagre brain can't work with that.

Even the column manufacturers will ask for samples to select the most suitable column if you plan on purchasing a prep column. They can predict the likelihood of success based on the enantiomer structure, but they often want to confirm it before you make an outlay on a very expensive column, usually about 3x the price of a similar-sized high-quality reversed phase column. Such columns are not as robust as NP and RP silica-based columns.

Bruce Hamilton

The basic mechanism in separation of chiral compounds is much more complicated then reverse phase, yes, the intermolecules forces like hydrogen bonding Pi interaction dipole moment and cavity creation etc is exectlly same but those can not separate the racemic into enantiomer, these only help to elute the same from column. only the stationary phase selectivity is main cause of separation of racemic into its enantiomer. I am working in separation of chiral compound from last 5 years I got some wired experience which theory can not explain.secoundly you can used some strong solvent like dichloromethank or DMSO in your mobile of some of Regis tech column.

Remember, in achiral separations, "one point" of contact is required for retentio/separation.

In Chiral separations, a "three point" contact is required for chiral discrimination.

This alone, makes the mechanisms more complicated and less predictable.

Wayne, seems that I have seen something like that before. Where does this one point versus three point stuff come from? I can´t think of any molecular interaction that could be operating only in chiral molecules.

All of the interactions are the same as discussed above, h-bonding, ionic, pi-pi, etc.

To recognize the difference between enantiomers, the CSP and analyte must interact at three points to discriminate. The three points, in different configurations on the differrent enantiomers, provide the separation. Remember, both enantiomers have the same functional groups for interantion, just in diferent spacial configurations, hence the need for a three point contact with stationary phase.

Hope that helps!

Sorry, dosn´t help, you are just restating your first statement. No refs?

Several models for the requirements to obtain chiral recognition have been proposed. The most reliable model is the three point contact model, peoposed by Dalgliesh (ref) that states that three interactions have to take effect and at least one of them has to be stereoselective.

ref. Dalgiesh, C.E. (1952), J. Chem. Soc. 137, 3940-3942.

My gut feeling: Forget about this. (I am, as it is not easy to go into the dungeon archives of our library).

Here is a brandnew reference of stuff like this:

The docking of chiral epoxides on the Whelk-O1 stationary phase: A molecular dynamics study • ARTICLE
Journal of Chromatography, In Press, Accepted Manuscript, Available online 23 March 2007,
Chunfeng Zhao and N.M. Cann

Really, any Chiral separations book will give an intro into the three point contact model. I just noted the original reference.

Thanks!

Found the article, "flew" over the abstract and conclusion and, though he uses the three point contact, he also indicates that he has evidence that the third point is complex (how many points is the third point?). So this just confirms my previous impressions that this 3 point thing is a construct that is used for calcs. etc., just like the Donnan membrane where there is no membrane, or the two layer ion thing. I didn´t see any new type of interactions.